Quote from: k on October 25, 2022, 03:49:52May be allegation is true. But one thing i will bring to notice. Unlike yester years AMD processors like athlon which worked much better than contemporary intels (atleast when compared with prices) on workstation like heavy computation jobs, may be Euler 3D or Specviewperf. Current AMD processors lack far behind from intel, even massive cored threadripper can't keep up with 12th and 13th gen Intel. This is serious and I relate this to memory bandwidth and (may be slightly to higher L1 cache of earlier AMD processors). AMD which was earlier having consistently higher memory bandwidth compared to intel counterpart has dumped that strategy and is using double channel even on 32 cores, is crazy beyond limits. And that is not from userbenchmark but consistent result from many sites. May be due to this reason though processor which is having high crunching power in general and performs good on software for which either processor is optimised or vice versa, will certainly lack on atleast few benchmarks. Overall ryzen may be best for gamers but is never cheap alternative for working professional/researchers for scientific simulations etc for xeon/Epyc, like athlon used to be for operaton etc. where it was great value. That honour goes to intel 12th gen and 13th gen now.   
So average poor score is what it deserve and not a bias.

Threadripper isn't Epyc. It's just Threadripper.

And amd has completely ignored that department by not releasing the non-pro versions.

Anyway, if they allow quad-channel there, then it would be competing with Epyc, thus reducing their margins.
You could call it the company's greed.
Now that we have 16 cores in Ryzen 9 series itself, threadripper would also need to be upgraded.
Same goes for Epyc.

Anyway, Threadripper is not a successor of Opteron.
It's competing with what used to be the HEDT segment of intel Core X processors.
They haven't been updated for a long time if I am not mistaken.

And even considering your point, userbenchmark is still s*** because of the way they compare processors with bias rather than basis lol

May be allegation is true. But one thing i will bring to notice. Unlike yester years AMD processors like athlon which worked much better than contemporary intels (atleast when compared with prices) on workstation like heavy computation jobs, may be Euler 3D or Specviewperf. Current AMD processors lack far behind from intel, even massive cored threadripper can't keep up with 12th and 13th gen Intel. This is serious and I relate this to memory bandwidth and (may be slightly to higher L1 cache of earlier AMD processors). AMD which was earlier having consistently higher memory bandwidth compared to intel counterpart has dumped that strategy and is using double channel even on 32 cores, is crazy beyond limits. And that is not from userbenchmark but consistent result from many sites. May be due to this reason though processor which is having high crunching power in general and performs good on software for which either processor is optimised or vice versa, will certainly lack on atleast few benchmarks. Overall ryzen may be best for gamers but is never cheap alternative for working professional/researchers for scientific simulations etc for xeon/Epyc, like athlon used to be for operaton etc. where it was great value. That honour goes to intel 12th gen and 13th gen now.   
So average poor score is what it deserve and not a bias.

Userbenchmark is and always has been a joke in tech circles, so this comes at no surprise whatsoever...

Quote from: DMC on October 24, 2022, 18:41:26

Quote from: bruh on October 24, 2022, 15:21:02

Quote from: SyCoREAPER on October 24, 2022, 05:19:47This sounded more like a fan fanboy rant than an a subjective analysis. You provided very little factual information buried inside the bashing and complaining.


And no I'm not an Intel boy. I have an AMD Advantage Laptop and Intel/Nvidia desktop so I have no bias when I say what I did above and that it could have been written a lot less ranty.

Nice tone policing! I couldn't POSSIBLY be mad about someone who deliberately lies to the same group my news website tries to cater to!

I wouldn't call it tone policing per se. While I see the article's overall point as legitimate, it's unlikely that bias is solely the cause of the skewed benchmarks. Unless the author is accusing UserBenchmark of flat out faking the results in favor of Intel, it should be stated that the Intel CPUs do perform better on certain workloads which is what leads to the perceived UserBenchmark bias. Whether those workloads are properly weighted in the results based on typical/practical use, and whether UserBenchmark was targeting these specific workloads for rational reasons are legitimate points for the author to question.

In general the article seems to imply implicit bias as the reason for the skewed results. While UserBenchmark is likely biased, it is also likely the bias is grounded in tangible performance superiority on specific workloads. The article should have focused on validity of those specific biases and in what use cases it misleads consumers to maintain a wholly objective approach.

It's pretty difficult to argue that there isn't an implict bias when they refer to AMD as "Advanced Marketing Devices."

Quote from: joe on October 24, 2022, 19:34:42

Quote from: Russell on October 24, 2022, 19:03:07Userbench has always been, still is and would always remain less than a meme.
But the point made in this article is very much relevant.
It is still a website that can actually influence the decision making of a potential buyer who doesn't know much about processors.
For those who are trying to talk about how 'intel processors are better in some instances'.
You can just read the articles on userbenchmarks showing skewed results. There's only one pattern that you'd observe, always trying to favour intel.

Even this same article here shows, that 7950X is faster than 12700K in 1-core and multi-core Userbenchmark tests. Potential buyer sees that clearly when comparing these cpu-s on Userbenchmark website. I dont see problem here unless benchmark is intentionally designed to perform better on some cpu-s.

The potential buyer who is a layman would most likely see this as i7 12xxx is better than latest Ryzen even though the later is new, so better go for i7 13xxx.

People are complaining that the easily decodable amd's new nomenclature is confusing.

Userbenchmark rated i7 11700k higher than Ryzen 9 5950x. That s*** was referred to as a massive waste of sand... It was even slower than it's previous gen i7 in many occasions...

Potential buyers who happen to be not related to tech would only ever look for the overall ranking or highest performace. Which is why amd, intel and nvidia are trying too hard by letting their TDPs run wild without inhibition.

If they were at least proficient to some degree, they'd at least check trustpilot or something and stay away from userbenchmark...

Quote from: Russell on October 24, 2022, 19:03:07Userbench has always been, still is and would always remain less than a meme.
But the point made in this article is very much relevant.
It is still a website that can actually influence the decision making of a potential buyer who doesn't know much about processors.
For those who are trying to talk about how 'intel processors are better in some instances'.
You can just read the articles on userbenchmarks showing skewed results. There's only one pattern that you'd observe, always trying to favour intel.

Even this same article here shows, that 7950X is faster than 12700K in 1-core and multi-core Userbenchmark tests. Potential buyer sees that clearly when comparing these cpu-s on Userbenchmark website. I dont see problem here unless benchmark is intentionally designed to perform better on some cpu-s.

Userbench has always been, still is and would always remain less than a meme.
But the point made in this article is very much relevant.
It is still a website that can actually influence the decision making of a potential buyer who doesn't know much about processors.
For those who are trying to talk about how 'intel processors are better in some instances'.
You can just read the articles on userbenchmarks showing skewed results. There's only one pattern that you'd observe, always trying to favour intel.

Quote from: bruh on October 24, 2022, 15:21:02

Quote from: SyCoREAPER on October 24, 2022, 05:19:47This sounded more like a fan fanboy rant than an a subjective analysis. You provided very little factual information buried inside the bashing and complaining.


And no I'm not an Intel boy. I have an AMD Advantage Laptop and Intel/Nvidia desktop so I have no bias when I say what I did above and that it could have been written a lot less ranty.

Nice tone policing! I couldn't POSSIBLY be mad about someone who deliberately lies to the same group my news website tries to cater to!

I wouldn't call it tone policing per se. While I see the article's overall point as legitimate, it's unlikely that bias is solely the cause of the skewed benchmarks. Unless the author is accusing UserBenchmark of flat out faking the results in favor of Intel, it should be stated that the Intel CPUs do perform better on certain workloads which is what leads to the perceived UserBenchmark bias. Whether those workloads are properly weighted in the results based on typical/practical use, and whether UserBenchmark was targeting these specific workloads for rational reasons are legitimate points for the author to question.

In general the article seems to imply implicit bias as the reason for the skewed results. While UserBenchmark is likely biased, it is also likely the bias is grounded in tangible performance superiority on specific workloads. The article should have focused on validity of those specific biases and in what use cases it misleads consumers to maintain a wholly objective approach.

Maybe they put 12700K in higher place, because its much cheaper than 7950X. You no, for consumer to know, that 12700K is better value.
I mostly look Userbenchmark, because there is also 2-core, 4-core, 8-core tests and easy to compare stuff. Other benchmarks only show 1-core and all-core results.
I noticed ranks only after fanboys started to fight about them.

Quote from: SyCoREAPER on October 24, 2022, 05:19:47This sounded more like a fan fanboy rant than an a subjective analysis. You provided very little factual information buried inside the bashing and complaining.


And no I'm not an Intel boy. I have an AMD Advantage Laptop and Intel/Nvidia desktop so I have no bias when I say what I did above and that it could have been written a lot less ranty.

Nice tone policing! I couldn't POSSIBLY be mad about someone who deliberately lies to the same group my news website tries to cater to!

@SyCoREAPER OK it's a rant. So what? You're no slouch in that department, either, as far as I can see.

This sounded more like a fan fanboy rant than an a subjective analysis. You provided very little factual information buried inside the bashing and complaining.


And no I'm not an Intel boy. I have an AMD Advantage Laptop and Intel/Nvidia desktop so I have no bias when I say what I did above and that it could have been written a lot less ranty.